Introduction To Ad Hoc Networks and Routing Protocols Presented By : Karthik Samudram Jayaraman

1
Introduction To Ad Hoc Networks and
Routing Protocols Presented By Karthik
Samudram Jayaraman
2
Introduction

• What is Ad Hoc Network?

• In Latin, ad hoc means "for this," further
  meaning "for this purpose only.
• All nodes are mobile and can be connected
  dynamically in an arbitrary manner.
• No default router available.
• Potentially every node becomes a router must be
  able to forward traffic on behalf of others.

3
Two types of wireless networks

• Infrastructured network
• A network with fixed and wired gateways. When a
  mobile unit goes out of range of one base
  station, it connects with new base station.
• Infrastructureless (ad hoc) networks
• All nodes of these networks behave as routers
  and take part in discovery and maintenance of
  routes to other nodes.

4
Why is Ad Hoc hard?

• Because of a constantly changing set of nodes
  Routing is a major issue!
• Security
• new vulnerabilities, nasty neighbors.
• Power
• running with batteries, little computing power.

5
The Desired Properties of Protocols (For Routing)

1. A routing protocol should be distributed.
2. Assume routes as unidirectional links.
3. Power efficient.
4. Consider its security.
5. Hybrid protocols can be preferred.

6
Two Main categories of Protocols

• Table Driven Routing Protocols
• Pro-active, learn the networks topology before
  a forwarding request comes in
• On-Demand Routing Protocols
• Re-active, become active only when needed

7
Destination Sequenced Distance Vector Algorithm
(DSDV)

• Based on Bellman-Ford Next Hop Routing.
• Each Node Maintains Tables for
• Next Hop on Path
• Distance (in hops) to destination.
• Sequence Number ( keep current route)

8

1. Nodes Exchange Updates With Neighbours

• Full Routing Updates
• Incremental Updates

DSDV contd
9
ClusterHead Gateway Switch Routing

• Adds hierarchical structure having DSDV as
• underlying routing algorithm.
• Routing is performed over clusterheads and not
• individual nodes.
• Requires Table to maintain cluster Membership, in
  
• addition to other routing tables.
• Cluster heads are selected by a Least Cluster
• Change Algorithm to minimize routing changes

10
Wireless Routing Protocol

• Each node maintains a distance table, a routing
  table, a link-cost table and a message
  retransmission list.
• Distance table of node i (matrix)
• For each destination j and each neighbor of i(k)
• Distance to j

11
Wireless Routing Protocol

• Routing table of node i is a vector
• The destinations identifier
• The distance to the destination
• The predecessor and successor of the chosen
  shortest path

12
Wireless Routing Protocol

• Link-cost Table
• The cost of relaying information through each
  neighbor
• Message retransmission list
• One or more retransmission entries

13
Wireless Routing Protocol

• Information Exchanged among nodes
• (routing table update messages )
• Identifier of the sending node
• A sequence number assigned by the sending node
• An update list of updates or ACKs to update
  message
• A response list of nodes that should send an ACK
  to the update message

14
Wireless Routing Protocol

• Each node will communicate with its neighbors
  reporting any changes in the system
• Each node will keep track of which node should
  send an acknowledgement
• Nodes will keep track of the changes in the
  system by periodic transmission of hello
  messages
• This protocol will force nodes to do consistent
  check of their predecessor hence avoiding
  count-to-infinity problem.

15
DSDV

• DSDV is based on idea of classical Bellman-Ford
  Routing Algorithm
• Each node maintains a routing table listing
  all available destinations. The attributes of
  each destination are the next hop, the number of
  hops to reach to the destination, and a sequence
  number, which is originated by the destination
  node.
• Both periodic and triggered routing updates to
  maintain table

16
Problems of Distance Vector

• Pro-active routing based on Distance Vector
• Topology changes are slowly propagated
• Count-to-infinity problem
• Moving nodes create confusion
• they carry connectivity data which are wrong at
  new place
• Table exchange eats bandwidth

17
DSDV

• How DSDV addresses the problems?
• Tagging of distance information
• The destination issues increasing sequence number
• Other nodes can discard old/duplicate updates
• Changes are not immediately propagated
• Wait some setting time
• Incremental updates instead of full table exchange

18
What is on-demand

• The routes are created when required
• The source has to discover a route to the
  destination.
• The source and intermediate nodes have to
  maintain a route as long as it is used.
• Routes have to be repaired in case of topology
  changes.

19
On-Demand Routing Protocols

1. Ad hoc On-demand Distance Vector Routing
2. Dynamic Source Routing Protocol
3. Temporally Ordered Routing Algorithm
4. Associativity Based Routing
5. Signal Stability Routing

20
Ad Hoc On-demand Distance Vector Routing

• AODV includes route discovery and route
  maintenance.
• AODV minimizes the number of broadcasts by
  creating routes on-demand
• AODV uses only symmetric links because the route
  reply packet follows the reverse path of route
  request packet.
• AODV uses hello messages to know its neighbors
  and to ensure symmetic links.

21
The source broadcasts a route packet
The neighbors in turn broadcast the packet till
it reaches the destination
source
RREQ
destination
RREP
Reply packet follows the reverse path of route
request packet recorded in broadcast packet
The node discards the packets having been seen
22
Route Maintenance

• If the source node moves, it reinitiates the
  route discovery.
• If intermediate node moves, its upstream node
  sends a RREP to the source. The source restarts
  the route discovery.

23
Dynamic Source Routing Protocol

• A node maintains route caches containing the
  routes it knows.
• Include route discovery and route maintenance.

24
Route discovery

• The source sends a broadcast packet which
  contains source address, destination address,
  request id and path.
• If a host saw the packet before, discards it.
• Otherwise, the route looks up its route caches
  to look for a route to destination, If not find,
  appends its address into the packet, rebroadcast,
• If finds a route in its route cache, sends a
  route reply packet, which is sent to the source
  by route cache or the route discovery.

25
source broadcasts a packet containing address of
source and destination
source
(1,4)
1
4
The destination sends a reply packet to source.
destination
8
(1,3)
3
7
(1,4,7)
2
The node discards the packets having been seen
(1,2)
6
5
(1,3,5,6)
(1,3,5)
The route looks up its route caches to look for a
route to destination If not find, appends its
address into the packet
26
How to send a reply packet

• If the destination has a route to the source in
  its route cache, use it
• Else if symmetric links are supported, use the
  reverse of route record
• Else if symmetric links are not supported, the
  destination initiates route discovery to source

27
Route maintenance

• Whenever a node transmits a data packet, a route
  reply, or a route error, it must verify that the
  next hop correctly receives the packet.
• If not, the node must send a route error to the
  node responsible for generating this route header
• The source restart the route discovery

28
Add entries into route cache

• The Source and destination in route discovery
• Intermediate hosts in route discovery
• The hosts receiving any broadcast

29
Temporally Order Routing Algorithm

• Creating Routes query/reply
• QRY packet is flooded through network
• UPD packet propagates back if route exist
• Maintaining Routes link-reversal
• UPD packets re-orient the route structure
• Erasing Routes
• CLR packet is floodthrough network to erase
  invalid routes

30
The source broadcasts a QRY packet with
height(D)0, all others NULL
(-,-,-,-,a)
(0,0,0,3,a)
(-,-,-,-,d)
(0,0,0,2,d)
QRY
a
source
d
Only the non-NULL node (destination) responds
with a UPD packet.
Dest.
h
QRY
QRY
(-,-,-,-,c)
(0,0,0,4,c)
UPD
(0,0,0,0,h)
c
g
(-,-,-,-,g)
(0,0,0,1,g)
b
(-,-,-,-,b)
(0,0,0,4,b)
f
e
(-,-,-,-,f)
(0,0,0,2,f)
(-,,-,-,-e)
(0,0,0,3,e)
A node receiving a UPD sets its height to one
more than UPD
Source receives a UPD with less height
31
TORA Height metric

• Each node contains a quintuple
• Logical time of a link failure
• Unique ID of the node that defined the new
  reference level
• Reflection indicator bit
• A propagation ordering parameter, height
• Unique ID of the node

32
Route Maintenance and Erasing

• No reaction necessary if all nodes still have
  downstream links.
• A new reference level is defined if a node loses
  its last downstream link.
• Synchronized clock is important, accomplished via
  GPS or algorithm such as Network Time Protocol.
• CLR packet to be flooded to clear the invalid
  packet.

33
Link failure with no reaction
(0,0,0,3,a)
(0,0,0,2,d)
a
d
Dest.
h
(0,0,0,0,h)
(0,0,0,4,c)
c
g
(0,0,0,1,g)
b
(0,0,0,4,b)
f
e
(0,0,0,2,f)
(0,0,0,3,e)
34
Re-establishing route after link failure
A new reference level is defined
(1,d,0,-2,s)
(0,0,0,4,s)
(1,d,0,-1,a)
(0,0,0,3,a)
(0,0,0,2,d)
(1,d,0,0,d)
UDP
s
a
d
UDP
Dest.
h
(0,0,0,0,h)
(0,0,0,4,c)
c
g
(0,0,0,1,g)
b
(0,0,0,4,b)
f
e
(0,0,0,2,f)
(0,0,0,3,e)
35
Associativity Based Routing

• Each route keeps a associativity table
• A high value of associativity tick indicates a
  low state of node mobility
• A route is selected based on associativity states
  of nodes, finds the high value of associativity
  tick (low mobility routes)

36
Associativity table

• All nodes generate periodic beacons
• When a neighbor node receives a beacon, it
  increases its associativity tick with respect to
  the sending node in associativity table
• Associativity ticks are reset when the neighbors
  of a node or the node itself move out of proximity

37
Route Discovery

• The source broadcast a QRY message
• Each intermediate node appends its address and
  associativity ticks to QRY,
• The destination can examine the associativity
  ticks to select route. If the multiple paths have
  the same overall degree of stability, select the
  minimum number of hops

38
Route Erasing

• If the the route is no longer desired, the source
  may not be aware of any route node changes
  because partial reconstruction.
• The source node initiates a route delete (RD)
  broadcast to erase the invalid route.

39
Conclusion
40

• DSR has lower routing load than AODV
• Because AODV has to depend on route discovery
  more often, DSR limits the overhead by using
  route cache
• TORA is higher because its overhead is the sum of
  neighbor discovery plus routing creating and
  maintenance

41

• DSR and AODV perform well than TORA, delivering
  over 95 packet regardless of mobility rate.
• TORA is lower because the link-reversal process
  fails in the routing maintenance.
• TORA has a better performance in the less sources.

42
Advantage and Disadvantage

• The overhead of TORA is worst. It has a better
  delivery ratio in less sources.
• DSR is good at all mobility rate and movement
  speed. Its performance is poor in a higher load.
• AODV performs almost as well as DST at all
  mobility rates and movement. It depends more on
  route discovery which may increase overhead in
• network

43
Overview
44
Reference

1. Routing Protocols for Ad Hoc Mobile Wireless
   Networt by Padmini Misra, ftp//ftp.netlab.ohio-s
   tate.edu/pub/jain/courses/cis788-99/adhoc_routing/
   index.htmlCBRP
2. A Comparison of On-Demand and Table Driven
   Routing for Ad-Hoc Wireless Networks, by Jyoti
   Raju  and J.J. Garcia-Luna-Aceves,
   http//www.soe.ucsc.edu/jyoti/paper2/
3. A New Routing Protocol for the Reconfigurable
   Wireless Networks, Zygmunt J Hass
4. Caching strategies in on-demand routing protocols
   for wireless ad hoc networks, by Yih-chun hu and
   Divid B. Johnson, http//monarch.cs.cmu.edu
5. Highly Dynamic Destination-Sequenced
   Distance-Vector Routing for Mobile Computers,
   Pravin Bhagwat, Charles E. Perkins
6. Dynamic source routing in ad hoc wireless
   networks, by David B. Johnson and David A. Maltz,
   http//www1.ics.uci.edu/atm/adhoc/paper-collectio
   n/johnson-dsr.pdf
7. A Performace Comparison of Multi-Hop Wireless Ad
   Hoc Network Routing Protocols, Josh Broch etc
8. An Efficient Routing Protocol for Wireless
   Netwrok, Shree Murthy etc
9. Temporally-Ordered Routing Algorithm (TORA)
   Version 1 Funtional Specification,  by V. Park,
   S. Corson, http//www1.ics.uci.edu/atm/adhoc/pape
   r-collection/corson-draft-ietf-manet-tora-spec-00.
   txt
10. Ad Hoc On Demand Distance Vector (AODV) Routing,
    by Charles Perkins, http//www1.ics.uci.edu/atm/a
    dhoc/paper-collection/perkins-draft-ietf-manet-aod
    v-00.txt

45
Reference (cont.)

1. An Introduction to Mobile Ad Hoc Network, by
   Ming Yu Jiang, http//kiki.ee.ntu.edu.tw/mmnet1/ad
   hoc/
2. Scalable Routing Strategies for Ad hoc Wireless
   Network, by Atsushi Iwata , Ching-Chuan Chiang
   etc.
3. A Performance Comparison of Multi-Hop Wireless Ad
   Hoc Network Routing Protocols, by Josh Broch,
   David A. Maltz, David B. Johnson, Yih-Chun Hu,
   Jorjeta Jetcheva, http//www1.ics.uci.edu/atm/adh
   oc/paper-collection/johnson-performance-comparison
   -mobicom98.pdf
4. Fisheye State Routing A Routing Schema for Ad
   Hoc Wireless Networks, by guangyu Pei, Mario
   Gerla, Tsi-Wei Chen
5. A review of current Routing protocols for ad-hoc
   Mobile Wireless Networks, by Elizabeth M. Royer
   and C-K Toh http//www.cs.ucsb.edu/vigna/courses/
   CS595_Fall01/royer99review.pdf
6. CEDAR a Core-Extraction distributed Ad Hoc
   Routing Algorithm, Prasun Sinha, Vaduvur
   Nharghavan, etc
7. Mobile computing today in the future, by M.J.
   Fahham and M.K. Hauge. http//www.doc.ic.ac.uk/n
   d/surprise_95/journal/vol4/mjf/report.html
8. Performance Comparison of On-demand Routing
   Protocols in Ad Hoc Network by Sohela Kaniz
   http//fiddle.visc.vt.edu/courses/ecpe6504-wireles
   s/projects_spring2000/pres_kaniz.pdf